Computing devices may utilize a dedicated user interface for speech recognition functions such as converting speech into text. Speech recognition mistakes can interrupt continuity of the dictation process, a problem that may be particularly frustrating to a user of a mobile device with limited input controls. Conventional systems may fail to provide an intuitive and efficient interface for speech recognition functionality on a mobile device. Accordingly, among other needs, there is an existing need for a speech recognition interface that is powerful yet efficient and easy to use.